2108 J. Org. Chem., Vol. 41, No. 12,1976
Pakrashi, Chattopadhyay, and Chakravarty
2.5 g of 1,6-dinitrophenazine which was recrystallized from glacial acetic acid, mp 348-349 "C (litS7343 OC). The third yellow band contained a mixture of 1,6- and 1,Q-dinitrophenazine.It was rechromatographed yielding 0.5 g of 1,6-dinitrophenazine and 1.5 g of 1.9dinitrophenazine which was recrystallized from 50% aqueous acetic (EtOH) acid, mp 267-270 O C (lit.7 273 OC). 1,6-Dinitrophenazine,,A, 246,365 nm. 1,6-Diaminophenazine. 1,6-Dinitrophenazine (816 mg) was suspended in 230 ml of 90%aqueous acetic acid, according to the literature? At the boiling temperature 1.8g of zinc powder was added with stirring in small portions over a period of 1.5 h. At the end of this period the dark red solution turned brown. Another 300 mg of Zn powder was added over 5 min. The solution was filtered and diluted with 230 ml of water. Concentrated ammonium hydroxide was added to pH 8 and the red precipitate was allowed to form during 24 h at 0 "C. It was taken up in 100ml of 2% hydrochloric acid, boiled for 2 h, treated with a little Norit, and filtered. Neutralization with concentrated ammonium hydroxide gave 632 mg of a red precipitate which was recrystallized twice from ethanol: yield 473 mg (75%)of 1,6-diaminophenazine; mp 250 OC (lit? 245 "C), TLC (silica gel), chloroform/ methanol (9:1),Rf 0.55; A,, (EtOH) 242,290,357,377,512 nm. The electronic spectrum did not change upon addition of sodium sulfite or sodium borohydride. Also in warm Me2SO the spectrum remained the same.
Registry No.-2-Nitrodiphenylamine-6,2'-dicarboxylic acid, 58718-49-3; iodinin, 68-81-5;2-chloro-3-nitrobenzoicacid, 3970-35-2; anthranilic acid, 118-92-3.
References and Notes (1) U. Hollstein and L. G. Marshall, J. Org. Chem., 37, 3510 (1972). (2) R. J. Pugmire, D. M. Grant, M. J. Robins, and R. K. Robins, J. Am. Chem. SOC.,91, 6381 (1969). (3) J. T. Clerk, E. Pretsch, and S. Sternhell, "'3C-Kernresonanzspektrosco-
pie", Akademische Verlaggesellschaft,Frankfurt am Main, West Germany,
1973. U. Hollstein, J. Heferocycl. Chem., 5, 299 (1968). H. Mcllwain, J. Chem. SOC., 1704 (1937). D. L. Vivian, Nature (London), 178, 753 (1956). S. Maffei, Gazz. Chim. [tal., 84, 667 (1954). K. Isono, K. Anzai and S. Suzuki, J. Antiblot., Ser. A, 11, 264 (1959). "Organic Syntheses", Collect. Vol. IV, Wiley, New York, N.Y., 1963, p 42. E. Lellman, Justus Liebigs Ann. Chem., 228, 243 (1885). S.Ball, T. W. Goodwin, and R. A. Marton, Biochem. J., 42, 576 (1948). U. Hollstein and D. A. McCamey, J. Org. Chem., 38, 3415 (1973). H. Akaborl and M. Nakamura, J. Antibiot.,Ser. A, 12, 17 (1959). I. Yosioka and Y. Kidani, J. Pharm. Soc. Jpn., 72,847 (1952). M. E. Flood, R. B. Herbert, and F. G. Holliman,J. Chem. Soc., Perkin Trans. 1, 622 (1972). (16) Y. Kidani and H. Otomasu, Hoshi Yakka Daigaki Kiyo, 6,39 (1957); Chem. Abstr., 52, 1181d(1958). (17) N. N. Gerber, J. Heterocycl. Chem., 6, 297 (1968). (18) H. P. Sigg and A. Toth, Helv. Chim. Acta, 50, 716 (1967).
(4) (5) (6) (7) (8) (9) 10) 11) 12) 13) 14) 15)
Studies on 4-Quinazolinones. 8.' Mechanism of Chromic Acid Oxidation of Arborine2 Satyesh Chandra Pakrashi,* Subhagata C h a t t ~ p a d h y a yand , ~ Ajit Kumar Chakravarty Indian Institute of Experimental Medicine, Calcutta 700032, India Received August 26,1975 Plausible mechanisms of formation of 2 and 3 from 1 by chromic acid oxidation have been advanced. A novel hydride shift from the side chain to the nucleus, or protonation across C=C in the alkene form, has been envisaged in the conversion of 1 to 2. Oxidation of a number of 1,2- and 2,3-disubstituted and 2-monosubstituted alkyl/aryl derivatives of 4-quinazolinone has also been studied and the formation of various products explained.
It was earlier reported4 that arborine (1),5 the major 4quinazolinone alkaloid of G. arborea (Roxb.) DC. (Rutaceae), underwent facile oxidation on brief heating with chromic acid in glacial acetic acid t o glycorine (2) and glycosmicine (3), the co-occurring6 minor bases in 78 and 14% yields, respectively. In view of t h e biogenetic implications,' we have studied this reaction in more detail t o gain insight into the mechanism of the interesting formation of 2 in particular. 0
0
0
n
H
Me
-
-2
Ib
IO
I
Me
Me
3
I
The same reaction could also be brought about to the extent of 50% by heating with aqueous chromic acid, while chromic oxide in pyridine at room temperature afforded 3 and benzaldehyde. On the other hand, glycosminine (4)upon heating with chromic acid in glacial acetic acid furnished the expected 2-benzoyl-4-quinazolinone( 5 ) as the sole product while 2benzyl-3-methyl-4-quinazolinone(6) yielded predominantly t h e 2-benzoyl derivative (7) along with 8 and 9. 0
0
0
4
5
78%
9
17%
-8 Me
0
0
7 -
H 2%
-9
Conceivably, the oxidation of 1 t o 2 with chromic acid involves the benzylidene form of arborine (lb),which almost certainly5~8takes part in all the oxidative processes studied so far,4,5,9s10and the inductive effect of the N-1 alkyl group. Should the benzylic form l a be involved, glycosminine, definitely showns to exist in this form (4)only, would have furnished benzoyleneurea (13) as one of the products, contrary t o t h e observation. Furthermore, oxidative debenzylation of 6, though less favored, also indicated the participation of the lone pair of electrons on nitrogen. A plausible mechanism put forward in Scheme I involves an initial electrophilic addition of HZCr04 to the benzylidene double bond of lb, analogous to that suggestedll,l* for the oxidation of alkenes. Attachment of the reagent at the benzylic carbon (path a) followed successively by a hydride shift13 (path al), further attack by the oxidant, and cleavage of C-C bond could then form 2. On the
J. Org. Chem., Vol. 41, No. 12,1976
Chromic Acid Oxidation of Arborine
2109
Scheme I 0
OJ$L Me I
Me
1
I
O-Cr(OH)2 I
I
Path b H2Cr04
Path a 2 HCrOi
Path a(
the
Ib, R = P h
0-
0-
I
0
R
0
- Cr(OH)2 I 0-
HCr0;j
1
1
HCr0;
d T oMe
I
B
0-
I Me
2
L
other hand, attachment of the second chromate ion a t C-2 without prior hydrogen transfer (path a*) or initial addition of the reagent a t the same position (path b) followed by another chromate ion attack would lead t o 3. In any case, the reduction of chromium valency provides the driving force for the bond rupture. In order to substantiate the proposed mechanism, we extended the original reaction to various other alkyl/aryl substituted 4-quinazolinones. The results summarized in Table I show that while 1,2-dialkyl derivatives (10) furnished 2 and 3 in varying amounts, the 2-alkyl derivatives (11) afforded products depending on the nature of the substituents. On the other hand, the 2,3-dialkyl derivatives (12) showed properties of both 1,2-disubstituted (10) and 2-monosubstituted (1 1) derivatives, the products and their yields being dependent on the substituent at position 2. I t can be seen from Table I that 2 is the major product from 10a and 10b wherein the hydride shift is possible (path al), with its'maximum yield from 10b due to better stability of its alkene form. The principal product 3 from 1Oc and 10f with no available hydrogen for transfer could easily be explained through path a2 or b (Scheme I). On the other hand, the forced oxidative decarboxylation of the N-1 methyl group in 10d in ca. 60%yield provided additional support to the participation of the alkene form in the reaction. Again, while the formation of 2 from 1Oc and 10d in small amounts could plausibly involve methyl migration (with concomitant attack by chromate ion in case of 10d)15followed by oxidation of the resulting loa, it would be difficult to rationalize its formation from 10f assuming phenyl migration sidce the 1-methyl-2-phenyl derivative (log) did not provide 2 even under more drastic conditions. Besides, hydride shift in complete exclusion of methyl or phenyl transfer has been shown14in simpler systems like 2-phenylpropene and 2,2-diphenylethylene. The migration of methyl/phenyl in our system was also believed t o be very unfavorable, if not altogether impossible. Furthermore,
Table I. Chromic Acid Oxidation Products of Substituted 4-Quinazolinones
I Me
a , R:Me, g , R:Ph,
Compd
10a 10b 1oca 10da 10f= 2 lla llb llc lld lleb 1 If lli llj
-
-
12
II
10
I
b , R = E t , c, R = C H M e 2 , d , R = C M e 3 i e,RcH,f,R:CHPh2j
h, R = C O M e j
I,
R=C(OH)Me2,
I,
R=C(OH)Ph2
Products (%)
l l d (60)
l l h (27) l l i (46) l l j (4% 5 (IO), benzophenone (25) 5 (12h
benzophenone (13) 12a 12b
12h (31)
Reaction time 0.5 h against 3 h for the rest. time 48 h.
Reaction
none of the reaction paths discussed above could satisfactorily explain the 18%yiel'd of 3 from 10d. We, therefore, cdnsidered protonation prior to chromate
2110 J.Org. Chem., Vol. 41, No. 12, 1976
Pakrashi, Chattopadhyay, and Chakravarty Scheme I1
Me
I Me
Re
LI 0
ion attack also as a distinct possibility. Thus, protonation across the alkene double bond in 1Oc and 10f would lead to 2 while protonation a t N-3 would eventually yield 3 from all the compounds discussed above (Scheme 11). We also took into account the possibility of even partial conversion of 2 to 3, in view of the reported17J8 oxidation of l l e to 13 in 5% yield on prolonged (7 days) reflux. This,
- Cr03H
envisaged (Scheme 111)in the formation of 13 could be confirmed by bringing about the same transformation directly Scheme I11 0
0
OH
Cr
OH
H
13
however, appeared unlikely because no such conversion could be detected after 0.5 h of reaction required for the complete transformation of 1 with ca. 14%yield of 3. Nevertheless, the latter could be obtained from 2 in ca. 9% yield in 3 h. In any case, the more facile oxidation of 2 compared to 1 l e further supported the postulated inductive effect of the N-1 methyl group and the mechanism could be explained through initial protonation a t N-3 (path b, Scheme 11). Now, in order to distinguish between the alternative postulates, viz. hydride transfer (Scheme I) or protonation (Scheme 11),we performed the following experiments. Arborine ( l ) ,both unlabeled and deuterated (with MeOD in the presence of sodium) a t the benzylic methylene (do, dl, d2; m/e 250,251,252), on being treated separately with chromic acid in DOAc (using D20 for working up), afforded 2 and its deuterated product (glycorine-do,-dl; m/e 160,161). On the other hand, replacement of DOAc with HOAc yielded unlabeled 2. The results would apparently support the protonation postulate. However, labeled 1 itself when heated on a steam bath with HOAc alone for 0.5 h or 2 on treatment with DOAc under the same condition exchangedlg deuterium. These results did not allow us to arrive a t a definite conclusion. Nevertheless, as has already been discussed, none of the postulates alone could explain the formation of the products of all the 4-quinazolinone derivatives studied. We, therefore, believe that both the reaction pathways compete with each other. The relative contribution of each would conceivably depend on the stability of the intermediate carbonium ion and as such on the structure of the individual compounds used. I t now remains t o explain the oxidation products of Nunsubstituted 2-alkyl-4-quinazolinones (Table I). Formation of l l e from l l a in reasonable yield apparently proceeds through oxidative decarboxylation (cf. quinaldic acidI6) of the C-2 methyl group. 2-Ethyl-4-quinazolinone ( l l b ) , like its benzyl analogue (4), underwent expected oxidation to the 2-acetyl derivative (1lh) as the major product which by further oxidative decarboxylation could lead to l le. On the other hand, 2-isopropyl- ( l l c ) and 2-diphenylmethyl- (1If) 4-quinazolinones afforded benzoyleneurea (13) and the corresponding 2,-(1'-hydroxy) derivatives ( l l i and llj)20in fairly good yields. The intermediacy of the latter
Cr "0
oQ
from 1 l i and 11j under identical condition. The additional product, viz. 2-benzoyl-4-quinazolinone(5),21obtained in ca. 10%yield from l l f could presumably be derived also from the intermediate l l j by further oxidation, analogous to the conversion of triphenylcarbinol to benzophenone.22As expected, 2-tert-butyl-4-quinazolinone ( l l d ) did not undergo oxidation a t all. In conclusion, chromic acid oxidation, like metal hydride r e d ~ c t i o n , ~once ~ , ~ again ~ - ~ ~demonstrates the unpredictable and yet most fascinating course of reactions of the 4-quinazolinone system.
Experimental Sectionz6 General Procedure for Oxidation with Chromium Trioxide in Acetic Acid. The 4-quinazolinone derivatives were prepared by known methods. Each compound to be oxidized was dissolved in a minimum volume of glacial acetic acid and heated with 2 molar eyuiv of CrOs over a steam bath for the period specified in Table I. The reaction mixture was then cooled, diluted with water, basified with ammonia, and extracted with chloroform. Glycorine (2) could be quantitatively isolated from the aqueous part by evaporating it to dryness and extraction of the residue with hot chloroform. It was characterized as picrate: mp 249 "C dec. Benzoyleneurea (13), mp 342-343 "C, was isolated as a crystalline solid in most cases on dilution of the reaction mixture with water and cooling. In the cases of l l f and 1lj, however, it wab isolated from the aqueous part on cooling after solvent extraction. All other crude products from the chloroform extract were separated and purified through chromatography over silica gel or alumina (neutral) followed by crystallization. The yields are given in Table I and the physical data of the products other than those reported earlier are mentioned below. 2-Benzoyl-4-quinazolinone ( 5 ) : mp 179-181 O C ; ir 3130 sh, 3030, 1706, 1680,1600 sh, 1597 cm-l. Anal. Calcd for C15HloN202: C, 71.99; H, 4.03; N, 11.20. Found: C, 71.88; H, 3.98; N, 11.05. 2-Benzovl-3-methvl-4-auinazolinone (7): mp 137 "C; ir 1700 sh, 1675,1663; 1582 cm'l. Anal. Calcd for C16H12N202: C, 72.71; H, 4.47; N, 10.67. Found: C, 72.55; H, 4.56; N, 10.44. 2-Acetyl-3-methyl-4-quinazolinone(12h): mp 78-79 "C; ir 1712, 1665,1605,1583cm-l.
Chromic Acid Oxidation of Arborine Anal. Calcd f o r C11HloN202: C, 65.33; H, 4.98; N, 13.86. Found: C, 65.54; H, 5.08; N, 13.59. 2-(l'-Hydroxyisopropyl)-4-quinazolinone(1li): mp 148 O C ; ir 3350, 3230, 3100, 1660, 1606 cm-'; NMR 6 1.73 s (6 H, M e d , 4.9 br ( O H ) , 7.25-7.85 m (3 14, A r H ) , 8.28 d (1H, J = 7 Hz, C5H). Anal. Calcd f o r CllH12N202: C, 64.72; H, 5.93;N, 13.72. Found: C, 64.94; H, 6.06; N, 13.58. Synthesis of 2-( l'-Hydroxyisopropyl)-4-quinazolinone(11). A solution o f anthranilamide (0.92 g) in dry p y r i d i n e (2 m l ) was treated w i t h 2-bromoisobutyryl b r o m i d e (1.2 m l ) a t 0 "C a n d the reaction m i x t u r e was k e p t a t r o o m temperature for 12 h. Crushed ice was added t o it a n d t h e separated solid (1.34 g) was filtered, washed, a n d dried. On repeated crystallizations f r o m benzene-alcohol it afmp 191-192 "C. forded N-(a-brnmoisobutyryl)anthranilamide, Anal. Calcd for CllH13N202Br: C 46.35; H, 4.60; N, 9.83; Br, 28.04. Found: C, 46.45; H, 4.57; N, 9.65; Br, 28.00. N-(a-Bromoisobutyry1)anthranilamide(0.1 g) was dissolved in 50% aqueous methanol (5 m l ) and heated over a steam b a t h w i t h ammonia solution (5 m l ) until a l l t h e solvent was evaporated. T h e residue o n p u r i f i c a t i o n t h r o u g h chromatography a n d crystallization f r o m benzene-petroleum ether afforded 1li as colorless needles, mp 147-148 O C , identical (melting point, m i x t u r e m e l t i n g point, ir, T L C ) w i t h t h e oxidation p r o d u c t o f llc.
Acknowledgment. We are grateful to the Council of Scientific and Industrial Research, New Delhi, for a senior fellowship to one of us (S.C.) and to Dr. Nitya Nand, CDRI, Lucknow, for the NMR spectrum. Registry No.-1,6873-15-0; 2, 3476-68-4; 2 picrate, 58718-50-6; 3, 604-50-2; 5, 13182-47-3; 7, 13182-48-4; loa, 7471-65-0; lob, 10553-04-5; lOc, 58718-51-7; lOd, 58718-52-8; IOf, 59123-36-3; lla, 1769-24-0; llb, 3137-64-2; l l c , 13182-64-4; lle, 491-36-1; llf, 18963-84-3; lli, 52827-42-6; llj, 18963-82-1; 12a, 1769-25-1; 12b, 58718-53-9; 12h, 58735-56-1; 13,86-96-4; c h r o m i u m trioxide, 133382-0; anthranilamide, 88-68-6; 2-bromoiuobutyryl bromide, 563-76-8; N - ( a - b r o m o i s o b u t y r y l ) anthranilamide, 58718-54-0.
References and Notes (1) Part 7: S. C. Pakrashi and A. K. Chakravarty, J. 04.Chem., 39,3828 (1974). (2) To be considered also as part 37 of the Studies on Indian Medicinal Plants series. For part 36 see B. Achari, A. Pal, and S.C. Pakrashi, Tetrahedron Lett.,4275 (1975). (3) Department of Chemistry, Jadavpur University, Calcutta 700032, India. (4) S.C. Pakrashi, and J. Bhattacharyya, Tetrahedron,24, 1 (1968). (5) D.Chakravarti, R. N. Chakravarti, L. A. Cohen, B. Das Gupta, S.Datta. and H. K. Miller, Tetrahedron, 16, 224 (1961).
J . Org. Chem., Vol. 41, No. 12, 1976 2111 (6) S. C. Pakrashi, and J. Bhattacharyya, Ann. Biochem. Exp. Med., 23, 123 (1963). (7) S.C. Pakrashi, J. indlan Chem. SOC., 44, 887 (1967). (8) S.C. Pakrashi,J. Bhattacharyya,L. F. Johnson, and H. Budzikiewicz, Tetrahedron, 19, 1011 (1963). (9) A. Chatterjee and S. Ghosh Majumdar, J. Am. Chem. SOC.,75,4365 (1953). (IO) A. Chatterjee and S.Ghosh Majumdar, J. Am. Chem. SOC.,76,2459 (1954). (11) W. J. Hickinbottom and G.E. M. Moussa, J. Chem. Soc., 4195 (1957): G. E. M. Moussa, J. Appi. Chem., 12, 385 (1962). (12) M. A. Davis and W. J. Hickinbottom, J. Chem. Soc., 2205 (1958). (13) Though the hydride shift does not seem to be known in case of chromic acid oxidation in glacial acetic acid, it has been reported in sulfuric acid medium [W. J. Hickinbottom, D. Peteres, and D. G. M. Wood, J. Chem. Soc., 1360 (1955)] and also in case of oxidation with chromyl ch10ride.l~ (14) F. Freemann, R. H. DuBois, and N. J. Yamachika, Tetrahedron,25, 3441 (1969). (15) Another plausible mechanism of conversion of 10d to 2 via an ylide type intermediate (cf. conversion of quinaldic acid to quinolinelB)through expulsion of tert-butyl carbonium ion from an N-protonated species could be ruled out since 10d remained unaffected on heating with glacial acetic acid alone. .(16) E. S.Gould, "Mechanism and Structure in Organic Chemistry", Holt, Rinehart and Winston, New York, N. Y., 1962, p 349. (17) St. V. Niementowski, Ber., 29, 1356 (1896). (18) W. L. F. Armarego, Adv. Heterocycl. Chem., 1, 253 (1963). (19) Though the acid-catalyzed deuteration at C-2 of 4-quinazolinone system is itself unknown, the relative ease with which it occurred (ca. 65%) was more unexpected in view of the reported drastic condition used (e.g., by heating in a sealed tube at 218 OC with DCI) for the exchange of the a proton of pyridine [J. A. Zoltewicz, and C. L. Smith, J. Am. Chem. Soc., 89, 3358 (1967)]. (20) (a) Such oxidation of isopropyl group to the corresponding tertiary alcohol with chromic acid is known.20b,cSimilar oxidation of the P-diphenylmeth I derivative in the quinazoline series has also been demonstrated by us. (b) L. F. Fieser and J. Szmuszkovicz, J. Am. Chem. Soc.. 70,3352 (1948). (c) L. F. Fieser, ibid., 70, 3237 (1948). (21) S.C. Pakrashi and A. K. Chakravarty, indlan J. Chem., 11, 122 (1973). (22) K. B. Wiberg, "Oxidation in Organic Chemistry", Academic Press, New York, N.Y., 1965, p 87. (23) S.C. Pakrashi and A. K. Chakravarty, Chem. Commun., 1443 (1969). (24) S.C. Pakrashi, J. Bhattacharyya, and A. K. Chakravarty, lndian J. Chem., 9, 1220 (1971). (25) S.C. Pakrashi, and A. K. Chakravarty, J. Org. Chem., 37, 3143 (1972). (26) All melting points were determined in open capillaries and are uncorrected. The ir spectra were taken in Nujol mull in a Perkin-Elmer lnfracord spectrophotometer (Model 137). The NMR spectrum of 111was recorded in a 60-MHz Varian instrument in CDCI:, with Me4SIas internal standard. Silica gel or alumina (neutral) was used for column chromatography. Thin layer chromatography was done on 0.3 mm silica gel plates using ethyl acetate-formic acid-chloroform (4:1:5) as the solvent system and the spots were developed by iodine vapor. Microanalyses were done by Dr. R. D. Macdonald, Micro-Analytical Laboratory, university of Melbourne, Australia. Identity of the oxidation products were established by comparison (melting point, mixture melting point, ir, TLC) with the authentic specimens or with those prepared according to procedures reported in the literature.
J,
